[rescue] Total corporate madness (

Shannon shannon at widomaker.com
Sat Aug 2 22:40:08 CDT 2003

On Thu, Jul 31, 2003 at 09:13:48PM -0500, Jonathan C. Patschke wrote:
> On Thu, 31 Jul 2003, Shannon wrote:
> > What I can't figure out is why the floppy interface is still so
> > slow.  There is no reason for it.
> The PC floppy interface is designed for bit-banging the hardware on the
> other end.  It's slow by design so that the controller won't outpace the
> mechanics of the drive.  There is nearly NO logic in a PC floppy drive.
> I refer you to the Minix PC floppy driver sources. :)

Yeah... like I said, there is no reason for it to be this way.  What
bothers me is even interfaces without such problems, often still
implement horribly slow floppies.

> However, there were IDE floppies for a while, and they didn't suck.
> But, in modern times, it seems pretty pathetic to waste an IDE port (and
> slow down the other device on the chain) for 2M of removable storage.

Yeah, but they keep justifying their waste with "transistors are cheap"
all the time, and implementing a fast and probably compatible floppy
interface would likely take almost nothing by today's standards.
Even if it is yet another IDE channel.

Personally, I wish IDE and USB had never happened, and we used nothing
but SCSI and FireWire, or something better.

I kept hoping that serial ATA was where they would finally do something
better. Sigh...

> But someone might try to make a duplicate copy of _The_Matrix_Reloaded_
> per their fair-use rights in case the original ever gets scratched, and
> then they wouldn't have to buy a new one.  SOMEONE THINK OF THE POOR

(Suzanne Sommer's voice)

    For only $16K per hour, you too can keep a starving RIAA member
    from enduring the pain of being yachtless and...

UNIX/Perl/C/Pizza__________________________________shannon at widomaker.com

More information about the rescue mailing list